The PhilPapers Fallacy (Series Announcement)
People often appeal to the results of the 2020 PhilPapers survey as evidence for or against a particular philosophical position. While these results provide some evidence, such appeals often overstate the degree to which we should defer to what a majority of philosophers believe (or don’t believe). One of the more common claims one might see, for instance, is that since 62% of respondents endorse moral realism, this is good evidence that moral realism is true.
I enthusiastically support gathering data of this kind, and hope that there are further efforts to evaluate the present state of the field. However, while these results provide a glimpse into what philosophers believe, it is an open question whether or not, and to what extent, the proportion who endorse a particular position is an especially strong indication that the position in question is true.
Over the next few weeks I will be making a case that we should put little stock in the results of these surveys as evidence for the views in question. I call overstating the results of these studies as evidence for a particular philosophical position the PhilPapers Fallacy, and hope to discourage using uncritical appeals to what most philosophers think to pressure those with less common views into ceding to the majority.
Genuine progress in the field will not be facilitated by those sympathetic to a majority position pressuring others to conform merely because it is the majority position, or by direct appeals to the purported expertise of philosophers. The final arbiter in any philosophical dispute is the quality of the arguments and evidence, not the assent of a simple majority nor a direct appeal to the expertise of specialists.
Below, I present an abstract and a slightly longer synopsis for the upcoming series:
Abstract
Surveys that indicate most analytic philosophers endorse a particular philosophical position do not necessarily provide strong evidence that the position is true. The strength of such appeals depends on the degree to which analytic philosophy confers appropriate expertise on its proponents, and on a host of background assumptions about the degree to which their judgments are independent of one another and free of parochial and biasing influences that would render their conclusions unrepresentative of the rest of the world’s population under similar reflective conditions. Until and unless legitimate skepticism about analytic philosophy’s methods is overcome, we should put little stock in what most analytic philosophers think.
Synopsis
People often appeal to the proportion of philosophers who endorse a particular view in the PhilPapers survey as evidence that a given philosophical position is true. Such appeals are often overused or misused in ways that are epistemically suspect, e.g., to end conversations or imply that if you reject the majority view on the matter that you are much more likely to be mistaken, or that you’re arrogant for believing you’re correct but most experts aren’t.
That most respondents to the PhilPapers survey endorse a particular view is very weak evidence that the view is true. Almost everyone responding to the survey is an analytic philosopher, and the degree to which the convergence of their judgments provides strong evidence is contingent on, among other things, (a) the degree to which analytic philosophy confers the relevant kind of expertise and (b) the degree to which their judgments are independent of one another.
There is good reason to believe people trained in analytic philosophy represent an extremely narrow and highly unrepresentative subset of human thought, and there is little evidence that the judgments that develop as a result of studying analytic philosophy are reflective of how people from other populations, or people under different cultural, historical, and educational conditions, would think about the same issues (if they would think about those issues at all).
Since most philosophers responding to the 2020 PhilPapers survey come from WEIRD populations, most of them are psychological outliers with respect to most of the rest of humanity. Their idiosyncrasies are further reinforced by self-selection effects (those who pursue careers in philosophy are more similar to one another than two randomly selected members of the population they come from), a narrow education that focuses on a shared canon of predominantly WEIRD authors, and induction into an extremely insular academic subculture that serves to further reinforce the homogenization of the thinking of its members. As such, analytic philosophers are, psychologically speaking, outliers among outliers among outliers.
At present, there is little evidence or compelling theoretical basis for believing that human minds would converge on the same proportion of assent to particular philosophical issues as what we see in the 2020 PhilPapers survey results if they were surveyed under different counterfactual conditions.
There is also little evidence nor much in the way of a compelling case that analytic philosophy confers expertise at being correct about philosophical disputes. The presumption that the preponderance of analytic philosophers sharing the same view is evidence that the view is correct is predicated, at least in part, on the further presumption that the questions are legitimate and that mainstream analytic philosophical methods are a good way to resolve those questions. Both of these claims are subject to legitimate skepticism. Analytic philosophy is a subculture that inducts its members into an extremely idiosyncratic, narrow, and comparatively homogeneous way of thought that is utterly unlike how the rest of humanity thinks. It has little track record of success and little external corroborating evidence of its efficacy.
Critics are not, therefore, obliged to confer substantial evidential weight on the proportion of analytic philosophers who endorse a particular philosophical position. Resolving how much stock we should put in what most philosophers think rests, first and foremost, on resolution about the efficacy of their methods.
TABLE OF CONTENTS
Part 2: Relevant expertise matters
Part 4: Selection effects matter
Part 5: How PhilPapers respondents interpreted the survey questions matters
Part 7: Philosophical fashions matter